CN105072632A - Energy efficiency optimization method in MIMO distributed base station system - Google Patents

Abstract

The invention discloses an energy efficiency optimization method in a single-user MIMO distributed base station system. The energy efficiency optimization method comprises the following steps: firstly, constructing an energy efficiency optimization problem model under a given remote access point set, and transforming the problem model into three sub-problems, namely a rate maximization problem, an energy efficiency optimization problem without rate constraint and a power minimization problem under the rate constraint, for solution, thereby obtaining the optimal solution of a transmission covariance and the maximal energy efficiency; and then orderly bringing access points into an activated access point set in an order from small to large distances between a user and the access points, obtaining the maximal energy efficiency based on the energy rate optimization problem model and determining an access point set finally selected to transmit information to the user. The energy efficiency optimization method in the single-user MIMO distributed base station system is capable of solving the problem of base station transmission pre-coding design in the single-user MIMO distributed base station system and the problem of how to select appropriate distributed access points to maximize the energy efficiency of the system; besides, the energy efficiency optimization method is capable of quickly converging the optimal solution.

Description

A kind of optimized method of energy efficiency in MIMO distributed base station system

Technical field

The invention belongs to the networking technology area in mobile communication system, particularly relate to energy efficiency optimization problem in a kind of mobile communications network.

Background technology

As one of performance index wanted in mobile communications network design, spectrum efficiency is widely studied between decades in the past, and MIMO distributing antenna system can improve the spectrum efficiency of system greatly.But due to shortage of energy and greenhouse effect, energy efficiency becomes the new study hotspot of of mobile communications network.And the efficiency optimization problem in MIMO spaced antenna is a non-convex problem, be difficult to try to achieve its optimal solution in existing technology.

Summary of the invention

Goal of the invention: the object of the invention is to the deficiency existed for prior art, a kind of optimized method of energy efficiency in MIMO distributed base station system is provided, the method solves on the one hand and selects the problem of remote access point: select a part of remote access point to carry out information transmission, is not the mode adopting remote access point standard-sized sheet; On the other hand, when given activation remote access point set, give one and solve the optimized problem of efficiency.

Technical scheme: for achieving the above object, the technical scheme that the present invention takes is as follows:

A kind of optimized method of energy efficiency in MIMO distributed base station system, comprises the steps:

(1) given remote access point set is set up under energy efficiency optimization problem model, and described problem model is converted into three subproblems and solves, the optimal solution of the transmission covariance making user's efficiency maximum under being met remote port maximum transmit power restrictions and the constraint of user's minimum-rate and corresponding maximum power efficiency value, described three subproblems are: maximize user rate problem (P1) under meeting remote port maximum transmit power restrictions, efficiency problem (P2) is maximized under meeting remote port maximum transmit power restrictions, and meet user's minimum-rate constraint and remote port maximum transmit power restrictions under minimum power problem (P3),

(2) user is arranged to the distance of remote access point according to order from small to large, each being included in by access point minimum for distance users activates access point set, the method of step (1) is adopted to calculate corresponding energy efficiency based on the set of given activation access point, until the energy efficiency that the access point increased makes access point set corresponding reduces, access point set corresponding to the maximum power efficiency obtained is the access point set finally selecting to send information to user.

Particularly, described given remote access point set under energy efficiency optimization problem model can be expressed as

Wherein, represent and activate the channel matrix of remote port of access to user, R _minfor the minimum-rate demand of user, for the hardware total power consumption of system, for all activated access point is to the correlation matrix of the transmission signal of user, represent and meeting each remote access point maximum transmit power under constraints feasible solution set, a is set the total number of middle access point, be s _ithe antenna number of individual access point.

Further, based on above-mentioned energy efficiency optimization problem model, described three subproblems are expressed as:

The step that described three subproblems solve is comprised:

(1.1) the transmission covariance making user rate maximum under (P1) is met access point power constraint is solved if then enter step (1.2), otherwise problem intangibility;

(1.2) the transmission covariance making user's efficiency maximum under (P2) is met access point power constraint is solved if then namely be the optimal solution of energy efficiency optimization problem under given remote access point set; Otherwise enter step (1.3);

(1.3) Solve problems (P3) be met access point power and user rate constraint under make user's efficiency maximum transmission covariance be the optimal solution of energy efficiency optimization problem under given remote access point set.

Further, in described step (1.1), the method for Solve problems (P1) is: problem (P1) Optimized model is converted into Lagrangian dual function Optimized model, optimal solution be expressed as wherein, λ _ifor Lagrange factor, U is to matrix carry out the matrix of the characteristic vector composition that singular value decomposition obtains, Λ=diag{q ₁..., q _rfor on r sub-channels transmitted power composition diagonal matrix; Optimal solution is tried to achieve by iterative algorithm specifically comprise:

(1.1.1) iterations initial value n=0 is made, random generation Lagrange factor λ ⁽⁰⁾=[λ ₁..., λ _a] make matrix $B={\mathrm{\Σ}}_{i=1}^A{\mathrm{\λ}}_i{B}_i\≥0;$

(1.1.2) calculate user and send covariance

(1.1.3) subgradient step-length is calculated

(1.1.4) upgrade u ⁽ⁿ⁾=1/n, if convergence, then stop iteration, obtain be optimal solution otherwise enter step (1.1.2), ε is the convergence threshold of setting.

Further, in described step (1.2), the method for Solve problems (P2) is: problem (P2) Optimized model is converted into function optimal solution is tried to achieve by iterative algorithm specifically comprise:

(1.2.1) make iterations n=0, initial value η is set ⁽⁰⁾, make G (η ⁽⁰⁾)=0;

(1.2.2) method of described step (1.1.1) to (1.1.4) is adopted to calculate now step (1.1.1) is replaced by the matrix B in (1.1.4)

(1.2.3) upgrade if | η ⁽ⁿ⁺¹⁾-η ⁽ⁿ⁾|≤ε restrains, then stop iteration, obtain be optimal solution otherwise enter step (1.2.2).

Further, in described step (1.3), the step of Solve problems (P3) comprising:

(1.3.1) η is remembered ^*=η ⁽ⁿ⁺¹⁾, wherein η ⁽ⁿ⁺¹⁾for solving the result that (P2) algorithm produces, make iterations n=1, initialization u ^max=η ^*, u ^min=0, u ⁽¹⁾=(u ^max+ u ^min)/2;

(1.3.2) method of described step (1.1.1) to (1.1.4) is adopted to calculate now step (1.1.1) is replaced by the matrix B in (1.1.4) if then stop iteration; Otherwise, make u ^max=u ⁽ⁿ⁾, u ⁽ⁿ⁺¹⁾=(u ^max+ u ^min)/2, repeat step (1.3.2) and carry out iteration.

Further, specifically comprise in described step (2):

(2.1) initial energy efficiency EE is set _opt=0, problem can line flag flag=0, and optimum access point set is the access point number A=1 activated; By the distance d of access point to user _iaccording to the order arrangement increased progressively, i.e. d _{π (1)}≤ ... ≤ d _{π (I)}, π (i) representative is from the index of the access point close to user i-th, and I is the total number of remote access point;

(2.2) order activates access point set the channel matrix carrying out self-activation access point is

(2.3) given in situation, by adopting the method for described step (1.1) to (1.3) to solve efficiency maximization problems, obtaining optimal solution, if optimization problem intangibility, so making otherwise, make flag=1, calculate according to optimal solution

(2.4) if order otherwise, stop, obtain the access point set finally selecting to send information to user

(2.5) if A < is I, make A=A+1, jump to step (2.2); Otherwise enter step (2.6);

(2.6) if flag=0, order based on velocities solved maximization problems (P1), calculates corresponding and all access points are opened; Otherwise, stop.

Beneficial effect: compared with prior art, the present invention has solved the efficiency optimization problem of distributed base station system under user rate constraint and distributed access point power limitation condition, and a kind of novel access point selection scheme based on distance is proposed, method is simple, optimal solution can be rapidly converged to, and result is accurate.

Accompanying drawing explanation

Fig. 1 is the network structure of MIMO distributed base station system.

Embodiment

Below technical solution of the present invention is described in detail, but protection scope of the present invention is not limited to described embodiment.

The embodiment of the invention discloses a kind of optimized method of energy efficiency in MIMO distributed base station system.As shown in Figure 1,

We consider there be I remote access point in a descending single user distributed base station, and 1 user.When remote access point number is less than 6, the position of a jth access point is (rcos (2 π (j-1)/I)), rsin (2 π (j-1)/I)), j=1,2, I, wherein r=2Rsin (π/I)/(3 π/I), otherwise first remote access point is positioned at center of housing estate (0,0), other I-1 access point is positioned at (rcos (2 π (j-1)/(I-1))), rsin (2 π (j-1)/(I-1))), j=2,, I.

The optimized method of energy efficiency of the embodiment of the present invention mainly comprises two parts, first under the set of given activation access point, propose a kind of transmission precoding optimization design scheme of low complex degree, efficiency optimization problem is converted into three subproblems and carries out solving (speed maximization problems by the program, there is no efficiency optimization problem under rate constraint, have minimum power problem under rate constraint), by the transmission pre-coding scheme (namely all activated access point is to the optimal solution of the correlation matrix of the transmission signal of user) of iterative optimum; Then a kind of novel low complex degree distributed access point selection algorithm based on distance is proposed.Concrete, the method for the embodiment of the present invention mainly comprises the steps:

(1) given remote access point set under energy efficiency optimization problem can be expressed as

Wherein, represent and activate the channel matrix of remote port of access to user, R _minfor the minimum of this user and rate requirement, for the total number of remote access point activated, the hardware total power consumption of system for all activated antenna sum, be s _ithe antenna number of individual access point, p _cfor the radio frequency link power loss that every root antenna is corresponding, p ₀for the static link loss of each remote port of access, for all activated access point is to the correlation matrix of the transmission signal of user, represent and meet each remote access point maximum transmit power under constraint feasible solution set,

Above-mentioned energy efficiency optimization problem is converted into and solves following three auxiliary subproblems by we: meet that the restriction of remote port maximum transmit power is lower maximizes user rate problem (P1), meet remote port maximum transmit power restriction lower maximization efficiency optimization problem (P2), and user's minimum-rate retrains and minimum power problem (P3) under remote port maximum transmit power restrictions.Three described subproblems are expressed as:

So solve the set of given activation access point under efficiency optimization problem following steps can be divided into solve:

(1.1) the transmission covariance making user rate maximum under (P1) is met access point power constraint is solved if then enter step (1.2), otherwise problem intangibility;

(1.2) the transmission covariance making user's efficiency maximum under (P2) is met access point power constraint is solved if then namely be the optimal solution of energy efficiency optimization problem under given remote access point set; Otherwise enter step (1.3);

(1.3) Solve problems (P3) is met access point power and user rate retrains the transmission covariance making user's efficiency maximum be the optimal solution of energy efficiency optimization problem under given remote access point set.

(2) user is arranged to the distance of remote access point according to order from small to large, each being included in by access point minimum for distance users activates access point set, by the method energy efficiency of step (1), until the access point increased makes energy efficiency reduce, the access point set obtained is the access point set finally selecting to send information to user.

Above-mentioned steps comprises the step that three subproblems solve in (1):

(2.1) step of Solve problems (P1)

To matrix carry out singular value decomposition wherein wherein λ _ifor Lagrange factor, U is the matrix of characteristic vector composition, D=diag{d ₁..., d _rit is matrix characteristic value composition diagonal matrix, first matrix Λ=diag{q is defined ₁..., q _rbe the diagonal matrix that the transmitted power on r sub-channels forms, wherein each diagonal element is then the detailed step solved to go wrong (P1) below:

The first step, iterations initial value n=0, random generation Lagrange factor λ ⁽⁰⁾=[λ ₁..., λ _a] make matrix $B={\mathrm{\&Sigma;}}_{i=1}^A{\mathrm{\&lambda;}}_i{B}_i\&GreaterEqual;0;$

Second step, calculating user send covariance

3rd step, calculating subgradient step-length

4th step, renewal ${\mathrm{\&lambda;}}_i^{(n+1)}=\&lsqb;{\mathrm{\&lambda;}}_i^{\left(n\right)}-u^{\left(n\right)}{s}_i^{\left(n\right)}\&rsqb;,\&ForAll;i,$ U ⁽ⁿ⁾=1/n, if $|{\mathrm{\&lambda;}}_i^{\left(n\right)}-{\mathrm{\&lambda;}}_i^{(n+1)}|\&le;\mathrm{\&epsiv;},\&ForAll;i$ Convergence, then stop iteration, otherwise repeat second step.

(2.2) step of Solve problems (P2)

Defined function so the specific algorithm flow process of Solve problems (P2) is:

The first step, initial value η is set ⁽⁰⁾, make G (η ⁽⁰⁾)=0, iterations n=0;

Second step, to obtain with the algorithm of problem (P1) now matrix B is replaced by

3rd step, renewal if | η ⁽ⁿ⁺¹⁾-η ⁽ⁿ⁾|≤ε restrains, otherwise repeats second step.

(2.3) step of Solve problems (P3)

Note η ^*=η ⁽ⁿ⁺¹⁾, wherein η ⁽ⁿ⁺¹⁾for solving the result that (P2) algorithm produces.

So the concrete grammar of Solve problems (P3) is as follows:

The first step, initialization u ^max=η ^*, u ^min=0, u ⁽¹⁾=(u ^max+ u ^min)/2, now iterations n=1;

Algorithm for Solving G (the u of second step, use problem (P1) ⁽ⁿ⁾) obtain now matrix B is replaced by if then stop iteration; Otherwise, u ^max=u ⁽ⁿ⁾, u ⁽ⁿ⁺¹⁾=(u ^max+ u ^min)/2.Repeat second step.

Wherein the concrete grammar of step (2) is as follows:

Initial energy efficiency EE is set _opt=0, problem can line flag flag=0, and optimum access point set is the access point number A=1 activated.By the distance d of access point to user _iaccording to the order arrangement increased progressively, i.e. d _{π (1)}≤ ... ≤ d _{π (I)}, π (i) representative is from the index of the access point close to user i-th.

The first step, order the channel matrix carrying out self-activation access point is

Second step, given in situation, solve efficiency maximization problems by the method for step one.If optimization problem intangibility, so makes otherwise flag=1, calculates

If the 3rd step order otherwise, stop;

If the 4th step A < I, makes A=A+1, jumps to the first step; Otherwise carry out the 5th step;

If the 5th step flag=0, order based on velocities solved maximization problems (P1), calculates corresponding and all access points are opened; Otherwise, stop.

The above invention is only the preferred embodiment of the present invention; be noted that for those skilled in the art; under the premise without departing from the principles of the invention; can also make some improvements and modifications that it is expected to, these improvements and modifications also should be considered as protection scope of the present invention.

Claims (7)

1. the optimized method of energy efficiency in MIMO distributed base station system, is characterized in that, comprise the steps:

(1) given remote access point set is set up under energy efficiency optimization problem model, and described problem model is converted into three subproblems and solves, the optimal solution of the transmission covariance making user's efficiency maximum under being met remote port maximum transmit power restrictions and the constraint of user's minimum-rate and corresponding maximum power efficiency value, described three subproblems are: maximize user rate problem (P1) under meeting remote port maximum transmit power restrictions, efficiency problem (P2) is maximized under meeting remote port maximum transmit power restrictions, and meet user's minimum-rate constraint and remote port maximum transmit power restrictions under minimum power problem (P3),

(2) user is arranged to the distance of remote access point according to order from small to large, each being included in by access point minimum for distance users activates access point set, the method of step (1) is adopted to calculate corresponding energy efficiency based on the set of given activation access point, until the energy efficiency that the access point increased makes access point set corresponding reduces, access point set corresponding to the maximum power efficiency obtained is the access point set finally selecting to send information to user.

2. the optimized method of energy efficiency in MIMO distributed base station system according to claim 1, is characterized in that, described given remote access point set under energy efficiency optimization problem model representation be

Wherein, represent and activate the channel matrix of remote port of access to user, R _minfor the minimum-rate demand of user, for the hardware total power consumption of system, for all activated access point is to the correlation matrix of the transmission signal of user, represent and meeting each remote access point maximum transmit power under constraints feasible solution set, a is set the total number of middle access point, be s _ithe antenna number of individual access point.

3. the optimized method of energy efficiency in MIMO distributed base station system according to claim 2, is characterized in that, described three subproblems are expressed as:

The step that described three subproblems solve is comprised:

(1.1) the transmission covariance making user rate maximum under (P1) is met access point power constraint is solved if then enter step (1.2), otherwise problem intangibility;

(1.2) the transmission covariance making user's efficiency maximum under (P2) is met access point power constraint is solved if then namely be the optimal solution of energy efficiency optimization problem under given remote access point set; Otherwise enter step (1.3);

(1.3) Solve problems (P3) be met access point power and user rate constraint under make user's efficiency maximum transmission covariance be the optimal solution of energy efficiency optimization problem under given remote access point set.

4. the optimized method of energy efficiency in MIMO distributed base station system according to claim 3, it is characterized in that, in described step (1.1), the method for Solve problems (P1) is: problem (P1) Optimized model is converted into Lagrangian dual function Optimized model, optimal solution be expressed as wherein, λ _ifor Lagrange factor, U is to matrix carry out the matrix of the characteristic vector composition that singular value decomposition obtains, Λ=diag{q ₁..., q _rfor on r sub-channels transmitted power composition diagonal matrix; Optimal solution is tried to achieve by iterative algorithm specifically comprise:

(1.1.1) iterations initial value n=0 is made, random generation Lagrange factor λ ⁽⁰⁾=[λ ₁..., λ _a] make matrix $B={\mathrm{\&Sigma;}}_{i=1}^A{\mathrm{\&lambda;}}_i{B}_i\&GreaterEqual;0;$

(1.1.2) calculate user and send covariance

(1.1.3) subgradient step-length is calculated

(1.1.4) upgrade u ⁽ⁿ⁾=1/n, if convergence, then stop iteration, obtain be optimal solution otherwise enter step (1.1.2), ε is the convergence threshold of setting.

5. the optimized method of energy efficiency in MIMO distributed base station system according to claim 4, it is characterized in that, in described step (1.2), the method for Solve problems (P2) is: problem (P2) Optimized model is converted into function optimal solution is tried to achieve by iterative algorithm specifically comprise:

(1.2.1) make iterations n=0, initial value η is set ⁽⁰⁾, make G (η ⁽⁰⁾)=0;

(1.2.2) method of described step (1.1.1) to (1.1.4) is adopted to calculate now step (1.1.1) is replaced by the matrix B in (1.1.4)

(1.2.3) upgrade if | η ⁽ⁿ⁺¹⁾-η ⁽ⁿ⁾|≤ε restrains, then stop iteration, obtain be optimal solution otherwise enter step (1.2.2).

6. the optimized method of energy efficiency in MIMO distributed base station system according to claim 5, is characterized in that, in described step (1.3), the step of Solve problems (P3) comprising:

(1.3.1) η is remembered ^*=η ⁽ⁿ⁺¹⁾, wherein η ⁽ⁿ⁺¹⁾for solving the result that (P2) algorithm produces, make iterations n=1, initialization u ^max=η ^*, u ^min=0, u ⁽¹⁾=(u ^max+ u ^min)/2;

(1.3.2) method of described step (1.1.1) to (1.1.4) is adopted to calculate now step (1.1.1) is replaced by the matrix B in (1.1.4) if then stop iteration; Otherwise, make u ^max=u ⁽ⁿ⁾, u ⁽ⁿ⁺¹⁾=(u ^max+ u ^min)/2, repeat step (1.3.2) and carry out iteration.

7. the optimized method of energy efficiency in MIMO distributed base station system according to claim 3, is characterized in that, described step specifically comprises in (2):

(2.1) initial energy efficiency EE is set _opt=0, problem can line flag flag=0, and optimum access point set is the access point number A=1 activated; By the distance d of access point to user _iaccording to the order arrangement increased progressively, i.e. d _{π (1)}≤ ... ≤ d _{π (I)}, π (i) representative is from the index of the access point close to user i-th, and I is the total number of remote access point;

(2.2) order activates access point set the channel matrix carrying out self-activation access point is

(2.3) given in situation, by adopting the method for described step (1.1) to (1.3) to solve efficiency maximization problems, obtaining optimal solution, if optimization problem intangibility, so making otherwise, make flag=1, calculate according to optimal solution

(2.4) if order otherwise, stop, obtain the access point set finally selecting to send information to user

(2.5) if A < is I, make A=A+1, jump to step (2.2); Otherwise enter step (2.6);

(2.6) if flag=0, order based on velocities solved maximization problems (P1), calculates corresponding and all access points are opened; Otherwise, stop.